Sentinel Lab — Satellite & Terrain Analytics

Measures seasonal canopy vitality, agricultural ripening, and crop harvest across southern and central Sweden. The analysis contrasts midsummer peak growth (June/July) against late summer harvest (August/September) by calculating the Normalized Difference Vegetation Index (NDVI) on atmospheric-corrected Bottom-of-Atmosphere (BOA) surface reflectance.

Key Finding: Intensive agricultural parcels around Lake Tåkern and across Skåne show a sharp NDVI plunge from ~0.85 (dense vegetative canopy) down to ~0.25 post-harvest. On Öland's limestone grassland (Alvaret), the index reflects moisture stress and drought response. Cloud, cloud shadow, and saturated pixels are automatically masked using the Sentinel-2 Scene Classification Layer (SCL).

Lake Tåkern is one of Northern Europe's renowned bird reserves, characterized by shallow open waters framed by vast reed beds (Phragmites australis). This module benchmarks classical McFeeters NDWI against Xu's Modified MNDWI utilizing Short-Wave Infrared radiation (SWIR-1 B11).

Hydrological Insight: Classical NDWI frequently overestimates open water across moist soils and marsh reed fringes. MNDWI strongly attenuates soil and vegetation reflectance in the SWIR spectrum, delineating open water surfaces with high fidelity. The analysis isolates a seasonal water body surface expansion of +0.68 km², accompanied by an interactive client-side threshold slider.

The record heatwave and drought of July 2018 triggered the most severe modern wildfires in Sweden's history. This experiment utilizes pair-matched, cloud-free Sentinel-2 observations captured immediately before and after the disaster to classify burn severity according to USGS / Copernicus Emergency Management Service standards.

Quantitative Damage Mapping: By contrasting NIR (B08, 842 nm, strongly absorbed by charcoal and charred biomass) against SWIR-2 (B12, 2190 nm, strongly reflected by bare ash and scorched mineral soils), the engine classifies exactly 6 102 hectares of high- and moderate-severity burned coniferous boreal forest.

Mount Omberg rises dramatically as an isolated bedrock horst between the deep Lake Vättern tectonic graben and the flat sedimentary plains of Östergötland. This experiment derives analytical morphometric terrain models directly from Copernicus DEM: elevation, slope gradient, and analytical shaded relief (hillshade).

Morphometric Findings: Elevations span from 87.5 m (Lake Vättern waterline) up to 263.6 m a.s.l. (Hjässan summit, ~176 m relative relief). Horn's 3×3 convolution filter detects steep cliff gradients of up to 50.5° slope along the western fracture line towards Lake Vättern. Includes an interactive SVG elevation cross-section profile!

Unlike naive web solutions that execute untrusted Javascript via eval() or new Function(), GeoX Raster Lab implements a secure, sandboxed lexical tokenizer and recursive descent parser compiling mathematical expressions into Abstract Syntax Trees (AST).

Capabilities: Interactive band selector tokens (B02, B03, B04, B08, B11, B12), arithmetic operators (+, −, *, /), one-click preset formulas (NDVI, NDWI, MNDWI, NBR, SAVI), real-time syntax checking with informative error messages, and GPU/canvas rendering using custom color ramps.